PROJECT SUMMARY A combination of entorhinal, hippocampal, and prefrontal pathology has a pivotal role in most neurological and neurodegenerative diseases and in the emergence of memory impairments that are associated with these diseases. Despite the knowledge that these brain regions are particularly vulnerable, the diversity of neural computations within and across these brain regions is only beginning to be revealed. For example, a key function of the hippocampus and medial entorhinal cortex (mEC) is to bridge events that are discontinuous in time, and entorhinal and hippocampal cells that are sequentially active (?time cells?) have been proposed to be pivotal for memory retention over delay intervals of many seconds. In our previous work, we therefore investigated the firing patterns over the delay interval in a spatial working memory (WM) task. We unexpectedly found that hippocampal time cells were not a general mechanism for WM retention in hippocampus-dependent tasks. Rather, preliminary data indicate that information about past and future choices during the delay interval is evident during brief population bursts in hippocampus, while mEC cells may show memory-related activity that persists irrespective of brain state. We therefore hypothesize that memory retention does not require time-varying activity over the delay interval but is rather evident in sporadic population bursts in hippocampus and medial prefrontal cortex (mPFC) throughout the delay period and in firing patterns of entorhinal cells that provide continuity irrespective of brain state. To record activity during the delay period with varying brain states, we will ? within each animal ? use two variants of a spatial WM task, one with and one without forced running throughout the delay such that either theta or non-theta states are predominant. Aim 1 will focus on population bursts in hippocampus and determine whether they are only informative in non-theta states, as shown in our preliminary data, or a general mechanism across brain states. In addition, we will selectively interrupt hippocampal activity within the delay period to determine whether coding of future choices by population bursts and behavior are perturbed. Aim 2 will then perform recordings across deep and superficial layers and along the dorso-ventral axis of entorhinal cortex to identify how mEC contributes to WM. Finally, Aim 3 will carry out large-scale combined single-unit and LFP recordings in hippocampus and mPFC and in mEC and mPFC to reveal the coordination of mechanisms for WM retention across brain regions and brain states. Taken together, we will identify whether neuronal activity during population bursts is a general mechanism for memory retention over delay intervals irrespective of brain state, whether mEC supports WM retention with persistent activity, and how subregions of mPFC are coordinated with hippocampal and entorhinal subareas along the dorso-ventral axis. Identifying these memory computations will be the foundation for treatment approaches to restore memory functions when brain circuits deteriorate in neurological and neurodegenerative diseases.